Although risk factors for coronary heart disease (CHD)⁴ have been extensively studied in various populations (Rose 1989, Welty et al. 1995), there has been relatively little such research among the Navajo. Although the Navajo have had a low prevalence of CHD and related risk factors during much of the past century (Coulehan and Welty 1990, Darby et al. 1956, Fulmer and Roberts 1963, Gilbert 1955, Salsbury 1937, Sievers 1967), and continue to have low rates of CHD compared with other American Indians (U.S. Department of Health and Human Services 1994a, Welty and Coulehan 1993), the prevalence of various CHD risk factors , along with rates of hospitalization and mortality, has increased over the last few decades (Klain et al. 1988, Sievers and Fisher 1979).

These trends may be partially explained by a shift from a traditional to a more Westernized lifestyle with substantial behavioral and dietary changes (DeStefano et al. 1979, Sugarman et al. 1990a, 1990b and 1992). A number of studies have found that secular trends in various CHD risk factors are followed by subsequent increases in CHD incidence and mortality (Epstein 1989). Increases among the Navajo in CHD risk factors such as hypertension, obesity, adverse blood lipids and diabetes may indicate, therefore, that CHD incidence will increase substantially among the Navajo. The Navajo Health and Nutrition Survey is the first population-based study to examine the prevalence of various CHD risk factors and their interrelationships among tribal members.

MATERIALS AND METHODS

Anthropometric measurements, made with respondents wearing light clothing and no shoes (White et al. 1997a) according to the standard protocols of the National Health and Nutrition Examination Survey III (McDowell et al. 1985), included weight, height, circumferences (waist and hip) and skinfolds (triceps, suprailiac, and subscapular). Body mass index (BMI) was calculated (kg/m²) as a measure of relative weight; in several analyses, we used various data from 20- to 29-y-olds in the National Health and Nutrition Examination Survey II (Najjar and Rowland 1987) to categorize subjects as underweight (<15th percentile), overweight (85th-94th percentiles), and obese (95th percentile). BMI cutpoints for overweight were 27.3 kg/m² (women) and 27.8 kg/m² (men); for obese, they were 32.3 kg/m² (women) and 31.1 kg/m² (men). The ratio of the subscapular skinfold thickness to the triceps skinfold thickness (STR) was used as an index of truncal obesity; although this ratio was less strongly related to BMI than were the separate skinfolds, truncal obesity was associated with BMI (r  0.2).

Three blood pressure measurements were taken with the use of a random-digit sphygmomanometer (Percy et al. 1997). The mean of the last two measurements was used in all analyses except for 23 respondents for whom only the first two measurements were available. Respondents were classified as having hypertension if systolic blood pressure was 140 mm Hg, diastolic blood pressure was 90 mm Hg or they reported taking antihypertensive medications, as verified by the interviewer (Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure 1988).

All laboratory measurements were performed in a single laboratory (Corning Clinical Laboratories, El Paso, TX) that participates in the proficiency survey operated by the College of American Pathologists. Specimens were transported to the laboratory on a daily basis, and concentrations of serum total cholesterol (TC) and triglycerides were measured with a Technicon SMAC-III (Technicon Instrument, Tarrytown, NY); a Technicon RAXT was used for determinations of serum HDL cholesterol and glucose. Concentrations of LDL cholesterol were estimated using the Friedewald equation (Friedewald et al. 1972): LDL = TC  [HDL cholesterol + (triglycerides/5)].

Our analyses excluded glucose, triglyceride and LDL cholesterol concentrations for the 62 participants whose fasting status was unknown or who reported fasting for <10 h. An LDL cholesterol concentration was also not calculated for the 22 participants with triglyceride concentrations >400 mg/dL.

Subjects were classified as diabetic on the basis of on self-report (if confirmed by medication or chart review) or by a positive diagnosis from an oral glucose tolerance test (WHO 1985, Will et al. 1997). Diabetes was defined as a fasting glucose 140 mg/dL or 2-h glucose 200 mg/dL after a 75-g oral glucose challenge. Lipid levels were classified according to the criteria of the National Cholesterol Education Program (1991): borderline-high TC (200-239 mg/dL) and high TC (240 mg/dL); low HDL cholesterol (<35 mg/dL); borderline-high LDL cholesterol (130-159 mg/dL), and high LDL cholesterol (160 mg/dL). Levels of triglycerides between 200 and 399 mg/dL and 400 mg/dL were classified as borderline-high and high, respectively. We also compared subjects' cholesterol concentrations with the responses of subjects about being informed by a health professional that they had a high cholesterol level.

Smoking and physical activity were self-reported, the latter based on the following question: During the past month, did you participate in any physical activities or exercises such as running, basketball, softball, aerobics, walking, hunting, getting wood, swimming or other activity for exercise?

RESULTS

Table 1. Levels of lipids and other characteristics among the Navajo, by sex and age [View Table]

Current smoking was reported by 5% of women and 16% of men, and its prevalence decreased with age (Table 2); among 20- to 39-y-olds, 23% of men and 8% of women reported currently smoking cigarettes. Smoking intensity was very low, with the median number of daily cigarettes ranging from 2 to 5 in the various sex-age groups. More men than women (29 vs. 21%) reported currently using chewing tobacco or snuff, with the highest prevalence among 20- to 39-y-olds. Overweight (or obesity) was very prevalent (>60%) among women, but less so among men; among men 60 years of age, the prevalence of overweight and obesity was only 9%. Overall, 23% of men and 14% of women were hypertensive, but rates for those age 60 and older were 44 and 28%, respectively. The prevalence of diabetes mellitus (~20%) increased dramatically with age, reaching ~40% among the older participants. About 25% of subjects reported that they had had no physical activity or exercise during the previous month.

Table 2. Prevalence of coronary heart disease risk factors among the Navajo [View Table]

Approximately 40% of the participants had a TC level  200 mg/dL, and 10% had a TC level  240 mg/dL; among these latter subjects, more than half reported never having been informed by a health care provider that their cholesterol was elevated (data not shown). Forty-six percent of men had an LDL cholesterol level 130 mg/dL, as did 27% of women. (Because the LDL cholesterol level was not estimated for 22 persons with a triglyceride level >400 mg/dL, these estimates should be interpreted cautiously.) An HDL cholesterol level <35 mg/dL was seen in 19% of the men and 11% of the women, and high or borderline-high triglyceride levels were seen in 28% of men and 21% of women. Additional analyses indicated that about one fourth of men and one third of women had a triglyceride level above the sex- and age-specific 90th percentile (Lipid Research Clinics 1980).

In general, BMI was significantly and positively associated with serum lipids, blood pressure and diabetes (Table 3). Associations between BMI and levels of lipids were stronger for men than for women; levels of total cholesterol, for example, showed no association with BMI among women (r = 0.04). In contrast, diabetes was strongly related to BMI among women but not in men, and systolic and diastolic blood pressure levels were related similarly to BMI among men and women (r = 0.29-0.34). The importance of obesity was also assessed by contrasting the proportions of persons in the normal and obese BMI groups who had diabetes mellitus, hypertension or adverse levels of various lipids (LDL cholesterol  160 mg/dL, HDL cholesterol <35 mg/dL or triglycerides  400 mg/dL). After adjustment for age and sex, obese persons were about four times as likely as those with normal BMI to have one or more of these risk factors (data not shown).

Table 3. Levels of various risk factors among the Navajo, by sex and body mass index (BMI) [View Table]

Truncal obesity, as assessed by STR, was associated with levels of HDL cholesterol, triglycerides and total/HDL cholesterol among men even after controlling for BMI (Table 4). For example, levels of HDL cholesterol decreased from 46 mg/dL among men in the lowest quartile of STR to 37 mg/dL among those in the upper quartile. Among women, truncal obesity was not significantly associated with any of the risk factors after adjustment for age and BMI, but there was a weak association with diastolic blood pressure (r = 0.11, P = 0.11). Additional analyses, in which subscapular and triceps skinfolds were added to models using age and BMI to predict the various risk factors, agreed well with the results obtained using the STR (data not shown).

Table 4. Relation of the subscapular/triceps skinfold ratio (STR) to various risk factors among the Navajo [View Table]

We also classified LDL cholesterol according to National Cholesterol Education Program guidelines (Table 5). Compared with persons having an LDL cholesterol concentration <130 mg/dL, those with concentrations  160 mg/dL were older, were more often diabetic and had a higher ratio of total cholesterol/HDL cholesterol.

Table 5. Levels of various risk factors according to LDL cholesterol level cutpoints among the Navajo [View Table]

DISCUSSION

For much of the past century, CHD, diabetes, obesity, hypertension and related risk factors had been rare among the Navajo (Darby et al. 1956, Fulmer and Roberts 1963, Gilbert 1955, Salsbury 1937, Sievers and Fisher 1979); currently, their CHD hospitalization and mortality rates are among the lowest in Indian Health Service service areas (Welty and Coulehan 1993). Nevertheless, heart disease is the second leading cause of death among the Navajo and thus a major public health concern (U.S. Department of Health and Human Services 1994a). Furthermore, because the clinical expression of CHD may occur only after decades of atherogenesis resulting from adverse levels of various risk factors (Goldberg 1992), and because the prevalence of these characteristics appears to have increased over the last few decades (Hall et al. 1992, Sugarman et al. 1990a and 1990b), the CHD burden of the Navajo will likely continue to grow.

Diabetes mellitus and its associated complications have risen markedly in this population during the 20th century (Sugarman et al. 1990a), and having both diabetes and other risk factors is now common among the Navajo (Hoy et al. 1994). Because obese persons in our study were about four times as likely to have diabetes, hypertension or adverse levels of at least one lipid, it is possible the substantial increase in overweight is an underlying cause. Sugarman et al. (1990b) also attributed the rise in diabetes to increased obesity among the Navajo; the prevalence of overweight among Navajo children rose rapidly between the 1950s and the 1980s, and the weight of school-age Navajo boys and girls increased by 29 and 19%, respectively. Although the adverse metabolic and clinical outcomes associated with a truncal or upper-body distribution of fat are well known (Björntorp 1992), there are few studies of fat patterning among American Indians (Szathmary and Holt 1983). We found, however, that both BMI and truncal obesity were more strongly related to lipid levels among men than among women.

It is also likely that adverse changes in levels of various lipids have occurred over the last few decades. For example, Sugarman et al. (1992) found that 34% of Navajo men (n = 105) and 18% of Navajo women (n = 150) had a total cholesterol level >240 mg/dL, and it was concluded that the Navajo no longer had the low serum lipid levels reported by earlier investigators (Fulmer and Roberts 1963, Sievers 1968). In our study, about 40% of Navajo adults had a total cholesterol level >200 mg/dL and about 10% had a level >240 mg/dL; these proportions are fairly comparable to those in the general U.S. population (Lipid Research Clinics 1980). Substantial differences, however, in levels of triglycerides and HDL cholesterol were observed, with 11% of women having an HDL cholesterol level <35 mg/dL, approximately the 5th percentile in the U.S. population (Lipid Research Clinics 1980). These findings, along with the markedly elevated levels of triglycerides and low levels of HDL cholesterol (particularly among women) among the Navajo, suggest a future increase in CHD rates.

Hypertension appears to have been uncommon among the Navajo in the first half of the century (Sievers 1977) but to have increased in recent decades. Darby et al. (1956) found prevalence rates of hypertension (>140/90 mm Hg) of 4-7% in two Navajo communities. Later, DeStefano et al. (1979) reported that the prevalence of a diastolic blood pressure >90 mm Hg among 640 adults from two communities was about 25% among men and <10% among women. A comparable analysis of the current data yields similar prevalence estimates, suggesting that there has not been a large increase in hypertension among the Navajo since the 1970s (data not shown).

Cigarette smoking was historically thought to have been uncommon among the Navajo (Sievers and Cohen 1961) and we, too, found that relatively few women and older men smoke. However, 23% of 20- to 39-y-old men reported currently smoking cigarettes and a substantial portion of younger men and women reported using chewing tobacco or snuff in our survey. Our findings of higher smoking among younger individuals may be of particular concern because tobacco use may also be increasing among Navajo adolescents (Davis et al. 1995, Freedman et al. 1997) and because almost all first use of tobacco in the general population occurs before graduation from high school (U.S. Department of Health and Human Services 1994b). Because most tobacco cessation programs have had only limited success (U.S. Department of Health and Human Services 1994b), prevention of tobacco use, especially among adolescents and young adults, may be a high priority for the Navajo. In addition, because the use of chewing tobacco and snuff has been shown to lead to subsequent smoking among other groups (U.S. Department of Health and Human Services 1994b), it may be important to examine this relationship among Navajo as well.

Finally, although physical activity among the Navajo has not been well studied, our findings and those of others suggest that the Navajo have a much more sedentary lifestyle than they did as recently as the 1960s (Sugarman et al. 1992).

Limitations of this study include its reliance on self-reported data and the absence of direct measures of prevalent heart disease (such as EKG readings). In addition, the assessment of physical activity was relatively crude, consisting of responses to a single question. Still, the study has many strengths, including the use of systematic sampling, standardized laboratory methodologies and quality control procedures, thus producing reliable estimates of CHD risk factor levels. Furthermore, few population-based studies among the Navajo have included levels of HDL cholesterol and have been able to examine interrelationships among the various risk factors.

Our findings concerning the prevalence of many CHD risk factors suggest that its incidence and mortality among the Navajo are likely to increase in the future. Such a development would be very much counter to the declining trends for CHD mortality in the general U.S. population seen over the last 30 years (McGovern et al. 1996). Increases in the prevalence of CHD risk factors may be partly responsible for rising CHD rates already reported among the Navajo. For example, Klain et al. (1988) found that the age-adjusted hospital discharge rate for acute myocardial infarction among Navajo men more than doubled between the mid-1970s and 1980s (from 0.84 to 2.03 per 1000 persons). These increases were attributed to the cumulative influence of secular trends in risk factors (primarily diabetes and hypertension) on preceding cohorts of Navajo.

These findings can help to assess the potential effect of CHD risk factors among the Navajo and assist in resource allocation for prevention strategies. For example, the low awareness of high serum cholesterol status among persons with a level  240 mg/dL suggests a need to improve screening and education campaigns, and the high prevalence of risk factors among middle-aged men may indicate a need for multifaceted health promotion activities. It may be most important, however, to develop better methods to prevent the onset of obesity and promote lifetime body weight control. Because of the relatively high prevalence of overweight among Navajo adolescents (Freedman et al. 1997), primary obesity prevention among children and adolescents may also be an important strategy, both because obesity among children is related to overweight in adulthood (Dietz 1987) and because sustained weight reduction is difficult among adults (National Institutes of Health Technology Assessment Conference Panel 1993). Prevention programs targeting body weight control by improved diet and physical activity probably hold the greatest promise for prevention of CHD among the Navajo in the future.